“Intelligent” cameras are not allowed in public spaces, Cnil decides

Faced with the proliferation of “augmented” or “intelligent” video devices in certain cities, the opinion of the National Commission for Data Processing and Freedoms, whose job is to guarantee respect for the privacy of the French, was particularly expected. Because these cameras, combined with automatic analysis software for filmed images, are much more intrusive than traditional devices. For this reason, after public consultation, the Commission published its position on this technology on Tuesday: as the law is, these devices are not approved for use by public authorities, she says. A legal framework must be provided to prepare “red lines”concludes the CNIL.

Uses that raise questions

On paper, these devices can be used in public space, in different types of situations listed by Cnil: for “tracks or traces“of people, the”crowd measurement and use of metro platforms or a station“, that “detection of suspicious events” (gatherings, offenses such as prohibited parking, wrong-way traffic or illegal dumping of waste, for example) or “abandoned objects“, that “characterization of the people filmed (age group, gender, behavior etc.)“or”allows the identification or characterization of persons by data processing (…) (colorimetric characterization of the used clothing)“. So we talk aboutboth statistical and safe use.

Last June, the city of Nice, the cradle of French video surveillance in public spaces, announced the rollout of software based on artificial intelligence that “will enable an analysis of the hazardous behavior of eight types of mobility in cities”, as explained by the mayor, Christian Estrosi. With such a level of monitoring and processing of collected information, automated and permanent, this type of technology constitutes “new ethical and legal issues“, according to the CNIL.

To view this Twitter content, you must accept cookies Social network.

These cookies make it possible to share or respond directly to the social networks to which you are connected, or to integrate content originally posted on these social networks. They also allow social networks to use your visits to our websites and applications for personalization and advertising targeting.


An attack on fundamental freedoms

For Cnil, “automatic and permanent analysis” rhymes with much more important data collection than when a human was responsible for image processing: “These new video tools can thus lead to massive processing of personal data, sometimes including sensitive data. People are therefore no longer only filmed by cameras, but analyzed in an automated way in order to derive in a probabilistic way certain information that, if necessary, enables decision-making or concrete measures regarding them.

What are the risks? A risk of generalized analysis of citizens in public space and an attack on the right to privacy, freedom of movement and assembly. More generally to the many basic, individual and collective freedoms, according to Cnil.

However, she will emphasize that the risks are not the same depending on the use of technologies: “An “enhanced” video device installed in a subway station and intended to study the density of use of the station will pose few direct risks to the people present if the purpose of the device is only to produce an attendance rate or to adjust and smooth traffic in real-time according to volume.

Conversely, more secure devices with “for the purpose or effect of decision-making or consequences at the individual level“, poses a problem. There would be a potential violation of citizens’ rights in the context of the detection of an offense and the arrest of its author, and even in the context of advertising targeting based on an analysis of the gender and the presumed age of a person passing a sign.

Legitimate and illegitimate entities

To date, there is no “specific texts framing the use of ‘augmented’ video devices“. For statistical purposes, these cameras are “can already be deployed without specific supervision. Cnil cites several examples of situations where their use is considered legal: “A device to calculate crowds in a subway to show travelers the least crowded trains to take, provided the results are anonymous”or “the processing makes it possible to analyze the type of frequency of a shopping center on the basis of criteria (gender of people and age group for example) to then allow the manager to choose the most suitable ads“.

On the other hand for “detection and prosecution of violations“, the public authorities’ use of these technologies is not authorized by French law, analyzes Cnil. It specifies: “Before any implementation of this type of system, managers must have clearly defined the objectives pursued, which must be fixed, explicit and legitimate.

In theory, the use of such devices is also subject to compliance with the applicable regulations regarding the protection of personal data (RGPD and the Data Protection Act), which in particular allow “ensure people the opportunity to oppose the processing of their data“. Which indeed seems impossible, unless this right of opposition cannot be waived by the establishment of a specific legal framework.

The CNIL also insists on the need to legislate: “Only a specific law, adapted to the technical characteristics and the efforts in question, could possibly, after a democratic debate, decide on their legitimacy and, by setting minimum guarantees, ensure a balanced reconciliation between the goal of protecting public order and the need to protect fundamental rights and freedoms.

Leave a Comment